fix: guard configurable monitoring ratios

api/src/infrastructure-monitoring/infrastructure-alert-settings.service.spec.ts

@@ -10,6 +10,8 @@ describe('InfrastructureAlertSettingsService', () => {
     expect(rules).toContain('HostCpuUsageWarning');
     expect(rules).toContain('CmppGatewayQueueDelayedCritical');
     expect(rules).toContain('threshold: "120秒"');
+    expect(rules).toContain('redis_memory_max_bytes > 0');
+    expect(rules).toContain('sum(increase(cmpp_api_http_requests_total');
   });
 
   it('rejects unknown keys and warning thresholds that are not below critical', () => {

api/src/infrastructure-monitoring/infrastructure-alert-settings.service.ts

@@ -14,12 +14,12 @@ export const ALERT_THRESHOLD_DEFINITIONS = [
   { key: 'hostCpu', label: '主机 CPU 使用率', unit: '%', min: 1, max: 100, step: 1, warning: 80, critical: 90, expr: '100 - (avg by (instance) (rate(node_cpu_seconds_total{mode="idle"}[5m])) * 100)', names: ['HostCpuUsageWarning', 'HostCpuUsageCritical'], service: 'host', durations: ['10m', '5m'] },
   { key: 'hostMemory', label: '主机内存使用率', unit: '%', min: 1, max: 100, step: 1, warning: 85, critical: 95, expr: '(1 - node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes) * 100', names: ['HostMemoryUsageWarning', 'HostMemoryUsageCritical'], service: 'host', durations: ['10m', '5m'] },
   { key: 'hostDisk', label: '根磁盘使用率', unit: '%', min: 1, max: 100, step: 1, warning: 80, critical: 90, expr: '(1 - node_filesystem_avail_bytes{mountpoint="/",fstype!~"tmpfs|overlay|squashfs"} / node_filesystem_size_bytes{mountpoint="/",fstype!~"tmpfs|overlay|squashfs"}) * 100', names: ['HostRootDiskUsageWarning', 'HostRootDiskUsageCritical'], service: 'host', durations: ['15m', '5m'] },
-  { key: 'apiError', label: 'API 5xx 错误率', unit: '%', min: 0.1, max: 100, step: 0.1, warning: 1, critical: 5, expr: '100 * sum(rate(cmpp_api_http_requests_total{status=~"5.."}[5m])) / clamp_min(sum(rate(cmpp_api_http_requests_total[5m])), 0.001)', names: ['CmppApiHttpErrorRateWarning', 'CmppApiHttpErrorRateCritical'], service: 'api', durations: ['5m', '5m'] },
+  { key: 'apiError', label: 'API 5xx 错误率', unit: '%', min: 0.1, max: 100, step: 0.1, warning: 1, critical: 5, expr: '100 * sum(rate(cmpp_api_http_requests_total{status=~"5.."}[5m])) / clamp_min(sum(rate(cmpp_api_http_requests_total[5m])), 0.001)', guard: 'sum(increase(cmpp_api_http_requests_total{status=~"5.."}[5m])) >= 5', names: ['CmppApiHttpErrorRateWarning', 'CmppApiHttpErrorRateCritical'], service: 'api', durations: ['5m', '5m'] },
   { key: 'apiLatency', label: 'API P95 响应时间', unit: '秒', min: 0.1, max: 60, step: 0.1, warning: 1, critical: 3, expr: 'histogram_quantile(0.95, sum by (le) (rate(cmpp_api_http_request_duration_seconds_bucket[10m])))', names: ['CmppApiLatencyWarning', 'CmppApiLatencyCritical'], service: 'api', durations: ['10m', '5m'] },
   { key: 'apiEventLoop', label: 'API 事件循环 P99', unit: '秒', min: 0.01, max: 10, step: 0.01, warning: 0.2, critical: 1, expr: 'cmpp_api_nodejs_event_loop_lag_p99_seconds', names: ['CmppApiEventLoopLagWarning', 'CmppApiEventLoopLagCritical'], service: 'api', durations: ['10m', '5m'] },
   { key: 'gatewayQueue', label: 'Gateway 最旧 pending', unit: '秒', min: 1, max: 3600, step: 1, warning: 30, critical: 120, expr: 'cmpp_gateway_submit_queue_oldest_pending_age_seconds', names: ['CmppGatewayQueueDelayedWarning', 'CmppGatewayQueueDelayedCritical'], service: 'gateway', durations: ['2m', '2m'] },
   { key: 'postgresConnections', label: 'PostgreSQL 连接使用率', unit: '%', min: 1, max: 100, step: 1, warning: 70, critical: 85, expr: '100 * sum(pg_stat_activity_count) / clamp_min(max(pg_settings_max_connections), 1)', names: ['PostgresConnectionsWarning', 'PostgresConnectionsCritical'], service: 'postgresql', durations: ['10m', '5m'] },
-  { key: 'redisMemory', label: 'Redis 内存使用率', unit: '%', min: 1, max: 100, step: 1, warning: 70, critical: 85, expr: '100 * redis_memory_used_bytes / redis_memory_max_bytes', names: ['RedisMemoryWarning', 'RedisMemoryCritical'], service: 'redis', durations: ['10m', '5m'] },
+  { key: 'redisMemory', label: 'Redis 内存使用率', unit: '%', min: 1, max: 100, step: 1, warning: 70, critical: 85, expr: '100 * redis_memory_used_bytes / redis_memory_max_bytes', guard: 'redis_memory_max_bytes > 0', names: ['RedisMemoryWarning', 'RedisMemoryCritical'], service: 'redis', durations: ['10m', '5m'] },
   { key: 'minioCapacity', label: 'MinIO 容量使用率', unit: '%', min: 1, max: 100, step: 1, warning: 80, critical: 90, expr: '100 * (1 - minio_cluster_capacity_usable_free_bytes / minio_cluster_capacity_usable_total_bytes)', names: ['MinioCapacityWarning', 'MinioCapacityCritical'], service: 'minio', durations: ['15m', '5m'] },
 ] as const;
 
@@ -110,7 +110,9 @@ export class InfrastructureAlertSettingsService {
       const values = [pair.warning, pair.critical];
       for (let index = 0; index < 2; index += 1) {
         const isWarning = index === 0;
-        const expr = isWarning ? `(${definition.expr} > ${values[0]}) and (${definition.expr} <= ${values[1]})` : `${definition.expr} > ${values[1]}`;
+        // 低样本量与“未配置容量上限”必须继续作为固定保护条件，避免单次错误或除零结果触发伪告警。
+        const guard = 'guard' in definition ? ` and (${definition.guard})` : '';
+        const expr = isWarning ? `(${definition.expr} > ${values[0]}) and (${definition.expr} <= ${values[1]})${guard}` : `(${definition.expr} > ${values[1]})${guard}`;
         lines.push(`      - alert: ${definition.names[index]}`, `        expr: ${expr}`, `        for: ${definition.durations[index]}`, '        labels:', `          severity: ${isWarning ? 'warning' : 'critical'}`, `          service: ${definition.service}`, '        annotations:', `          summary: "${definition.label}${isWarning ? '达到警告阈值' : '达到严重阈值'}"`, `          description: "${definition.label}持续超过${values[index]}${definition.unit}。"`, '          currentValue: "{{ $value }}"', `          threshold: "${values[index]}${definition.unit}"`);
       }
     }

docs/testing-progress.md

@@ -3664,3 +3664,4 @@ git diff --check
 - 系统监控将“服务关键指标”移至活动告警正上方，标题说明明确使用 Prometheus；新增十组固定指标警告/严重阈值设置。配置真实写入 PostgreSQL，使用版本条件并发认领，promtool 校验、原子替换和回环 reload 成功后才推进生效版本，失败保留旧生效值。
 - 右上角预警中心新增“系统监控告警”及数量/严重数，读取独立 Prometheus 活动告警汇总并跳转活动告警锚点；继续使用 `Promise.allSettled` 隔离签名、安全与监控域故障。安全检测页移除重复大号标题，说明明确使用 Fail2ban。
 - 新增 migration `20260814173000_add_infrastructure_alert_settings`。本机 Prisma Client 生成、前后端 TypeScript 与 API 正式编译已通过；测试机恢复资产、migration、Prometheus规则校验、真实告警数据、浏览器验收和提交/部署结果待本轮后续补记。
+- 首次测试机规则验收发现 Redis 未配置 `maxmemory` 时除零结果会误触发 Critical pending，立即补回 `redis_memory_max_bytes > 0` 固定保护，并补回 API 5xx 窗口至少 5 次错误的最低样本门槛；该真实验收缺陷未按通过处理。

tools/monitoring/cmpp-managed-alerts.yml

@@ -32,12 +32,12 @@ groups:
         labels: { severity: critical, service: host }
         annotations: { summary: "根磁盘使用率达到严重阈值", description: "根磁盘使用率持续超过90%。", currentValue: "{{ $value }}", threshold: "90%" }
       - alert: CmppApiHttpErrorRateWarning
-        expr: (100 * sum(rate(cmpp_api_http_requests_total{status=~"5.."}[5m])) / clamp_min(sum(rate(cmpp_api_http_requests_total[5m])), 0.001) > 1) and (100 * sum(rate(cmpp_api_http_requests_total{status=~"5.."}[5m])) / clamp_min(sum(rate(cmpp_api_http_requests_total[5m])), 0.001) <= 5)
+        expr: (100 * sum(rate(cmpp_api_http_requests_total{status=~"5.."}[5m])) / clamp_min(sum(rate(cmpp_api_http_requests_total[5m])), 0.001) > 1) and (100 * sum(rate(cmpp_api_http_requests_total{status=~"5.."}[5m])) / clamp_min(sum(rate(cmpp_api_http_requests_total[5m])), 0.001) <= 5) and (sum(increase(cmpp_api_http_requests_total{status=~"5.."}[5m])) >= 5)
         for: 5m
         labels: { severity: warning, service: api }
         annotations: { summary: "API 5xx 错误率达到警告阈值", description: "API 5xx 错误率持续超过1%。", currentValue: "{{ $value }}", threshold: "1%" }
       - alert: CmppApiHttpErrorRateCritical
-        expr: 100 * sum(rate(cmpp_api_http_requests_total{status=~"5.."}[5m])) / clamp_min(sum(rate(cmpp_api_http_requests_total[5m])), 0.001) > 5
+        expr: (100 * sum(rate(cmpp_api_http_requests_total{status=~"5.."}[5m])) / clamp_min(sum(rate(cmpp_api_http_requests_total[5m])), 0.001) > 5) and (sum(increase(cmpp_api_http_requests_total{status=~"5.."}[5m])) >= 5)
         for: 5m
         labels: { severity: critical, service: api }
         annotations: { summary: "API 5xx 错误率达到严重阈值", description: "API 5xx 错误率持续超过5%。", currentValue: "{{ $value }}", threshold: "5%" }
@@ -82,12 +82,12 @@ groups:
         labels: { severity: critical, service: postgresql }
         annotations: { summary: "PostgreSQL 连接使用率达到严重阈值", description: "PostgreSQL 连接使用率持续超过85%。", currentValue: "{{ $value }}", threshold: "85%" }
       - alert: RedisMemoryWarning
-        expr: (100 * redis_memory_used_bytes / redis_memory_max_bytes > 70) and (100 * redis_memory_used_bytes / redis_memory_max_bytes <= 85)
+        expr: (100 * redis_memory_used_bytes / redis_memory_max_bytes > 70) and (100 * redis_memory_used_bytes / redis_memory_max_bytes <= 85) and (redis_memory_max_bytes > 0)
         for: 10m
         labels: { severity: warning, service: redis }
         annotations: { summary: "Redis 内存使用率达到警告阈值", description: "Redis 内存使用率持续超过70%。", currentValue: "{{ $value }}", threshold: "70%" }
       - alert: RedisMemoryCritical
-        expr: 100 * redis_memory_used_bytes / redis_memory_max_bytes > 85
+        expr: (100 * redis_memory_used_bytes / redis_memory_max_bytes > 85) and (redis_memory_max_bytes > 0)
         for: 5m
         labels: { severity: critical, service: redis }
         annotations: { summary: "Redis 内存使用率达到严重阈值", description: "Redis 内存使用率持续超过85%。", currentValue: "{{ $value }}", threshold: "85%" }